THE JOURNAL OF Bro~ocrca~ CHEMISTRY Vol. 265, No. 17, Issue of June 15, pp. 10081-10087,199O 0 1990 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S. A.

Unfolding Intermediates in the Triple Helix to Coil Transition of Bovine Type XI Collagen and Human Type V Collagens &ar2 and cula2cr3*

(Received for publication, February 9, 1990)

Nicholas P. Morris$$, Sandra L. Watts, Janice M. Davis*, and Hans Peter BBchinger$$ll From the $Research Department, Shriners Hospital for Crippled Children and the SDepartment of Biochemistry and Molecular Biology, Oregon Health Sciences University, Portland, Oregon 97201

The thermal triple helix to coil transitions of two human type V collagens (a12a2 and ala2a3) and bovine type XI collagen differ from those of the interstitial collagens type I, II, and III by the presence of unfolding intermediates. The total transition enthalpy of these collagens is comparable to the transition enthalpy of the interstitial collagens with values of 17.9 kJ/mol tripeptide units for type XI collagen, 22.9 kJ/mol for type V (alza2), and 18.5 kJ/mol for type V (ala2a3). It is shown by optical rotatory dispersion and differ- ential scanning calorimetry that complex transition curves with stable intermediates exist. Type XI colla- gen has two main transitions at 38.5 and 41.5 “C and a smaller transition at 40.1 “C. Type V (al~a2) shows two main transitions at 38.2 and 42.9 “C and two smaller transitions at 40.1 and 41.3 “C. Compared to these two collagens type V (ala2a3) unfolds at a lower temperature with two main transitions at 36.4 and 38.1 “C and two minor transitions at 40.5 and 42.9 “C. The intermediates present at different temperatures are characterized by resistance to trypsin digestion, length measurements of the resistant fragments after rotary shadowing, and amino-terminal sequencing. One of the intermediate peptides has been identified as belonging to the (~2 type V chain, starting at position 430 and being about 380 residues long. (The residue numbering begins with the first residue of the first amino-terminal tripeptide unit of the main triple helix. The a2(XI) chain was assumed to be the same length as the al(XI).) One intermediate was identified from the a2(XI) chain and with starting position at residue 495, and three from the a3(XI) with starting positions at residues 519,585, and 618.

Collagens are a family of gene products with a common structural motif, the triple helix. This conformation consists of three chains with the sequence (Gly-X-Y), which form left- handed polyproline II helices. Three of these helices form a right-handed superhelix called a triple helix (1,Z). This struc- ture is stabilized by hydrogen bonds between the amino group of glycine and the carbonyl group of the X residue in a neighboring chain and possibly by additional hydrogen bonds involving solvent molecules (3). The X and Y positions are

* This work was supported by the Shriners Hospital for Crippled Children. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

ll To whom correspondence should be addressed: Research Dept., Shriners Hospital, 3101 SW. Sam Jackson Park Rd., Portland, OR 97201. Tel.: 503-241-5090 (ext. 543).

often occupied by proline. Proline residues in the Y position become hydroxylated in a post-translational modification (4). The hydroxyproline residues contribute significantly to the stability of the triple helix (3, 5, 6).

The interstitial collagens, types I, II, III, V, and XI, are a subgroup of the gene family which together form the extensive banded quarter-staggered collagen fiber systems characteris- tic of the vertebrate extracellular matrix. Recently it has been observed that these fibers are heterotypic. In cartilage and related tissues the fibers contain predominantly type II col- lagen and small amounts of type XI (7), while noncartilage tissues contain type I collagen as the abundant form with small amounts of type III and type V (8, 9).

Besides having a related function, these collagens share many other structural features including: a major uninter- rupted triple helical domain 300 nm in length containing about 1000 amino acids/chain, a high degree of sequence homology, and a conserved organization of introns and exons within the genome, a similar degree of prolyl hydroxylation, and nontriple helical domains at either end of the major triple helix which are proteolytically processed during fiber forma- tion, although the extent and rate of processing vary (10-16). The minor collagens, types V and XI, have additional distinc- tive features. Type V collagen is found in alternative forms (aLa2 and (~la2a3) depending on the tissue. Type XI collagen is composed of three different chains (ala2a3), two of which are unique (oil and a2), while the third (LYE) appears to be identical to the cul(I1) chain.’ The precise role of types V and XI collagen in the formation and function of the interstitial collagen fiber system is unknown.

The triple helices of interstitial collagens undergo sharp thermal transitions from triple helix to coil just above the body temperature of the species from which the collagen was isolated. From studies with collagens and collagen-like poly- tripeptides it is known that this transition is highly coopera- tive (3). In many cases including types I and II collagen, an all or none model describes this transition very well, indicat- ing the absence of unfolding intermediates (3, 17, 18). How- ever, interruptions in the Gly-X-Y sequence which occur in type IV collagen of basement membranes lead to independ- ently melting domains (19). In this report we present a de- scription of the thermal unfolding of type V collagens (al*~u2 and (~lot2a3)* and type XI collagen which do not contain interruptions but, in contrast to type I and type II collagens, show unfolding intermediates.

’ R. Glanville, personal communication. * Type V collagen with the chain composition nl,ot2 is referred to

as type V, and type V collagen consisting of ala2a3 as type V,.

10081

by guest on March 13, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Thermal Stability of Type XI and V Collagens

MATERIALS AND METHODS

Purification of Type V and XI Collagens-All purification proce- dures were carried out at 4 “C. Type V, collagen was purified from human amnion essentially as described (20) with minor changes. Briefly, amnion was separated from human placentas and homoge- nized in cold distilled water. The homogenate was centrifuged at 4000 rpm for 20 min in a RC-3 centrifuge, and the pellet was resuspended in cold 1 M NaCl. This material was centrifuged as above, and finally resuspended in cold distilled water and centrifuged again as above. The pellet was weighed and resuspended at about 250 g/liter in cold 0.5 M acetic acid with 1 g of pepsin (Worthington), 100 g of tissue pellet and digested for 18 h at 4 “C. The digest was then centrifuged (Sorvall RC-5B) at 13,000 X g., for 20 min, after which the superna- tant was taken to 10% (w/v) NaCl and stirred for 20 min. Differential salt fractionation was carried out as described (20) with sequential precipitation at 2.7 and 4 M NaCl at neutral pH. The 4 M NaCl precipitate was further fractionated at 4 and 8% (w/v) NaCl in 0.5 M acetic acid. The last precipitated (8% NaCl insoluble fraction) ma- terial, containing mostly type V collagen, was then suspended in and dialyzed against DEAE buffer (50 mM Tris/HCl, pH 8.3, containing 2 M urea and 0.2 M NaCl) and applied to a DEAE-cellulose column (Whatman) in the same buffer. The unbound fraction was pooled and precipitated by adding acetic acid to a final concentration of 0.5 M and 14% (w/v) NaCl, and the precipitate collected by centrifugation at 13,000 X g., for 40 min. The pellet was resuspended in a small volume of CM-cellulose buffer (0.04 M LiAc, pH 4.8, containing 0.05 M LiCl and 2 M urea) and dialyzed against CM-cellulose buffer. The material was applied to the CM-cellulose column and eluted using a linear gradient from 0.05 to 0.28 M LiCl. Type V collagens from pooIed fractions were precipitated and cohected as described above for DEAE-cellulose fractions. The final pellet was resuspended in 50 mM Tris/HCl buffer, pH 7.5, containing 400 mM NaCl and dialyzed against this buffer. Type V, collagen was purified from placenta as described above, and further separated from type V, collagen by a combination of ammonium sulfate precipitation (20) and dialysis into phosphate-buffered saline (21). Type XI collagen was purified as described previously (12, 22). All collagens were treated with diiso- propyl fluorophosphate (1 mM final concentration) to inhibit pro- teases.

Determination of Protein Concentration-Concentrations were de- termined by amino acid analysis (23) and by circular dichroism measurements using 7500 degree cm’/dmol as the molar elhpticity per residue at 221 nm.

Determination of Thermal Stability by Resistance to Trypsin Diges- tion-unfolding studies were performed on diisopropyl fluorophos- phate-treated purified collagen samples as previously described (12). loo-p1 aliquots of the samples (0.5-1.0 mg/ml) in 50 mM Tris/HCl buffer, pH 7.5, containing 400 mM NaCl were equilibrated at the indicated temperature for 10 min in a water bath, at which time 10 ~1 of a 200 rg/ml solution of L-l-tosylamido-2-phenylethyl chloro- methyl ketone-trypsin (Sigma) in the same buffer was added. Exactly 2 min afterwards, the digestion was stopped by the addition of 10 ~1 of 10 x SDS3 sample buffer (12) and heating in a hot oil bath (110 “C) briefly. Samples were then incubated at 95 “C for 4 min in a dry heating block.

Electron Microscopy-Samples for electron microscopy were treated as above, except that at the end of the 2-min digestion, a 25- ~1 aliquot of the sample was removed and added to 75 ~1 of 0.75 M acetic acid on ice. The acidified samples were dialyzed against 0.5 M acetic acid at 4 “C, and then prepared for rotary shadowing with platinum and carbon as described (23). Electron micrographs were then analyzed for the length distribution of molecules using the BioQuant (R&M Biometrics Inc., Nashville, TN) system.

Optical Rotatory Dispersion Measurements-The thermal stability was measured by ORD at 365 nm in a 241 MC polarimeter (Perkin- Elmer) equipped with a loo-mm thermostatted quartz cell. The temperature was controlled by a circulating water bath (RCS, Lauda Division, Brinkman Instruments) and was increased linearly by a programmable controller (PM 350, Lauda) at a rate of 10 or 1 ‘C/h. The temperature was monitored within the cell with a thermistor and a digital thermometer (Omega Engineering, Inc., Stamford, CT). The analog signals of both instruments were recorded and digitized on a XY-plotter (HP 7090A, Hewlett-Packard) and the data was stored

” The abbreviations used are: SDS, sodium dodecyl sulfate; ORD, optical rotatory dispersion; CAPS, 3-(cyclohexylamino)-l-propane- sulfonic acid.

on an IBM PC/XT computer (IBM Corp.). Differential Scanning Calorimetry-The temperature dependence

of the partial heat capacity was measured in a MC-2 differential scanning calorimeter (MicroCal Inc., Northampton, MA). Data col- lection and analysis were performed on an IBM PC/AT computer with the DA-2 software from MicroCal, according to published pro- cedures (24, 25).

Blotting and Amino Acid Sequence Determination of the Interme- diate Fragments-After electrophoresis of the fragments, the proteins were transferred to polyvinylidene difluoride (Immobilon, Millipore Corp.) membranes using the method described by Matsudaira (26). The buffer utilized was 10 mM CAPS, 0.1% EDTA, pH 11.0, with 10% methanol and transfer was accomplished at 4 “C. After transfer, membranes were stained with Coomassie Blue (26), air dried, and the bands of interest were cut out with a clean scalpel blade. Sequence determination was made on a model 470A sequenator equipped with a model 120A phenylthiodantoin analyzer (Applied Biosystems).

RESULTS

The thermal denaturation of type II collagen, as monitored by ORD shown in Fig. 1 (panel A), yields a single sharp transition with a transition midpoint (7’“) of 41 “C reflecting the highly cooperative nature of the unfolding process. There is no evidence of transition intermediates. However, when type V, is analyzed in the same way (Fig. 1, panel B), the triple helix to random coil transition is broader and more complex. There appear to be at least two transitions, one with a !F, similar to type II collagen, consisting of about 40% of the 0R.D signal, and one which occurs at a lower temperature with about 60% of the ORD signal. Types V, and XI collagens show a similar thermal denaturation pattern (Fig. 1, panels C and D, respectively). Again about 80 and 60%, respectively, unfold at the lower temperature transition with the remainder being more stable.

Analysis of the thermal denaturation by differential scan- ning calorimetry provides a measure of the transition enthal- pies as well as a more detailed view of the unfolding process (Fig. 2) and these are summarized in Table I. Again, type II collagen shows a single narrow transition (Fig. 2, panel A).

B ‘-; i

J

Temperature PC)

FIG. 1. Thermal unfolding of type II, type V,, type V,, and type XI collagen monitored by ORD. The thermal transitions were observed at 365 nm with a heating rate of 10 “C/h in 50 mM Tris/HCl buffer, pH 7.5, containing 400 MM NaCl. Panel A, bovine type II collagen (0.28 mg/ml); panel B, human type V, collagen (a12012) (0.26 mg/ml); panel C, human type V, collagen (ola2a3) (0.29 mg/ml); and panel D, bovine type XI collagen (1.2 mg/ml).

by guest on March 13, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Thermal Stability of Type XI and V Collagens 10083

25 30 36 40 45 50

Temperature Idee C)

FIG. 2. Differential scanning calorimetry of the unfolding transition of type II, type V,, type V,, and type XI collagen. The partial heat capacity was determined as a function of temperature in 50 mM Tris/HCl buffer, pH 7.5, containing 400 mM NaCl with a heating rate of 10 or 13 “C/h. Panel A, bovine type II collagen (1.0 mg/ml); panel B, human type V, collagen (011*(~2) (1.59 mg/ml);panel C, human type V, collagen (011012~3) (0.29 mg/ml); and panel D, bovine type XI collagen (0.49 mg/ml).

TABLE I TABLE I

Transition midpoints CT,) and transition enthalpies of the triple Transition midpoints CT,) and transition enthalpies of the triple helix coil transitions of bovine type II, bovine type XI, and human helix coil transitions of bovine type II, bovine type XI, and human

type V collagens (~~1~~72) and (ala2~~3,J type V collagens (~~1~~72) and (ala2~~3,J

Collagen type T”, AlF Total m

“C

Bovine type II Bovine type XI

Human type V, (&(~2)

Human type V, ((ula2a3)

42.8 38.5 40.1 41.5 38.2 40.1 41.3 42.9 36.4 38.1 40.5 42.9

kJ/mol tripeptide unit

25.5 9.1 3.2 5.6 17.9

10.9 2.1 2.0 7.9 22.9 7.0 7.2 1.6 2.7 18.5

Type V, collagen yields a complex transition profile (panel B) with two major and two minor peaks resolved by deconvolu- tion of the heat capacity as a function of temperature. The portion of type V, which unfolds at higher temperature has a T,,, similar to that of type II (42.8 “C) while the portion unfolding at lower temperature has a T, of about 38.2 “C. Type XI collagen has a very similar distribution of transition temperatures. Surprisingly, type V, has two predominant transitions at much lower temperatures, 38.1 and 36.4 “C, even though type V, shares two chains with type V,.

There are several possible explanations for the complex thermal denaturation profiies displayed by type V and XI collagens. (a) The rate of temperature increase used in these experiments was too rapid compared to the rate of denatura- tion thereby trapping kinetic intermediates. (b) Each type of collagen contained two or more populations of molecules, each behaving like type II collagen, but having different transition midpoints. (c) Types V and XI collagen contain triple helical domains of differing stability within the molecule.

To address the first point, the thermal denaturation of type V,, monitored by ORD, was repeated at a rate of heating of 1 “C/h and compared to the original experimental condition of 10 “C/h. As shown in Fig. 3, the slower rate of heating decreases the T,,, by about 1.5 “C, but the shape of the tran- sition was unaltered. At the slower heating rate, two or more transitions are present in the same relative amounts as shown in Fig. 2, panel B.

In order to distinguish between the second and third pos- sibilities mentioned above, the direct demonstration of the absence or presence of transition intermediates is required. If types V and XI collagen are composed of subpopulations of molecules each of which behaves like type II but with a different T,,,, then one would not expect to find transition intermediates at any temperature, but rather fully native or fully denatured molecules. However, if these collagens, which are uninterrupted triple helices, have regions or domains of different stability, then one would expect to find evidence of the more stable domain during the transition. This was ana- lyzed in two ways,

First, at each temperature of the transition, an aliquot was removed and digested with trypsin. The triple helix is resist- ant to proteolysis with trypsin, while the denatured chains are digested to very small peptides. The results were analyzed by SDS-polyacrylamide gel electrophoresis using 5-10% ac- rylamide gradient gels. As shown in Fig. 4, panel A, type II collagen shows only full sized a chains which disappear near the T,. The resulting small peptides run in the dye front and are not detectable. Type V, (Fig. 4, panel B) shows full sized (Y chains as the first transition approaches. As these disappear at around 38 “C a set of peptides with greater mobility appear and these in turn begin to disappear at the start of the second transition, 43 “C. Both type V, (panel C) and type XI (panel D) show similar behavior when analyzed by sensitivity to trypsin.

Second, the trypsin-resistant triple helical conformation was analyzed by rotary shadowing and electron microscopy. By this technique, the triple helix is observed as a distinct

10 20 30 40 50 60 70 Temperature (%)

FIG. 3. Thermal unfolding of type V, collagen as a function of the heating rate. The triple helix coil transition of type V, collagen (nl~a2) was measured by ORD at 365 nm in 50 mM Tris/ HCl buffer, pH 7.5, containing 400 mM NaCl at 10 and 1 “C/h. Lowering the heating rate by a factor of 10 shifts the denaturation temperature down by 1.5 degrees. However, the shape of the transition remains the same.

by guest on March 13, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Thermal Stability of Type XI and V Collagens

.#..ae-c,...-

1 5 10 15 19

FIG. 4. Thermal unfolding of type II, type V,, type V,, and type XI collagen as monitored by resistance to trypsin diges- tion. Aliquots of each type of collagen were incubated at various temperatures in 50 mM Tris/HCl buffer, pH 7.5, containing 400 mM NaCl for 10 min followed by a 2-min digestion with trypsin. 5-10% SDS-polyacrylamide gels are shown of the reduced samples after Coomassie Blue staining. For all panels the temperatures are: lane I, 25 “C; lane 2, 30 “C; and lanes 3-19, 1 “C above the temperature in the previous lane. Panel A, bovine type II collagen; panel B, human type V, collagen (nlrtu2); panel C, human type V, collagen (nla2a3); and panel D, bovine type XI collagen.

rope-like molecule while denatured chains are not detectable because they lack a sufficient profile to accumulate metal (19). Type V, collagen was incubated at 25, 38, or 40 “C, treated with trypsin as above, and rotary shadowed. At 25 “C, the molecule is completely resistant to trypsin and the full length image is seen by electron microscopy in Fig. 5, panel a. At 38 “C, type V is in the middle of the first transition and truncated, about half-length, molecules are observed (Fig. 5, panel b). At 40 “C type V collagen is near the middle of the second transition and truncated molecules somewhat shorter than those obtained at 38 “C are observed (Fig. 5, panel c). This experiment was also performed with type V, and type XI collagens. The lengths of the molecules at each tempera- ture were determined and plotted as a histogram in Fig. 6, confirming that the electron micrographs are representative. In each case there is an intermediate form of the molecule

FIG. 5. Electron micrographs of rotary-shadowed type V, collagen treated with trypsin at various temperatures. Trypsin digestion was performed in 50 mM Tris/HCl buffer, pH 7.4, containing 400 mM NaCl at 25 (panel a), 38 (panel b), and 40 “C (panel c). The bar indicates a length of 100 nm.

with a distinct length at the given temperature. These trun- cated triple helices correspond to the residual conformation detected by ORD and differential scanning calorimetry in the complex denaturation profiles shown above in Figs. 1 and 2.

Intermediate fragments were analyzed by blotting onto polyvinylidene difluoride membrane followed by amino-ter- minal sequencing. Fig. 7 depicts the polyacrylamide gels that were blotted onto polyvinylidene difluoride membrane and indicates the peptides that were sequenced. Table II shows a summary of sequences obtained and their identification of the region in the protein. Unfortunately not all sequences of the constituent O( chains of these collagens are known and we found only one peptide of the type V, digests that could be identified. The fragment starts at position 4304 of the a2 chain and is, based on an average length of 160 nm determined by electron microscopy, about 380 residues long. For type XI collagen one peptide matches a sequence in the (~2 chain (14). Three more peptides are perfect matches with sequences from the human otl(I1) chain, consistent with the idea that the a3(XI) chain is identical to the (ul(I1) chain.

DISCUSSION

The thermal stability of type V collagen has been reported previously (20, 27-33) with considerable variation among the results. Much of the discussion in these reports concerned the chain composition of the type V molecules. Data in three reports indicate the presence of multiple transitions, although this was not the main focus of these reports. Studies on the susceptibility of V, and V, to various proteases suggest that localized unfolding occurs before the molecules are completely denatured (29). Details of the temperature dependence of protease susceptibility were not addressed, however. In order to demonstrate the distinct properties of types V, and V,, Niyibizi et al. (20) show intermediate fragments produced by trypsin digestion at 34, 35, and 37 “C. These authors also report thermal transition curves monitored by circular di- chroism. Although the temperature range of the transitions is similar to that reported here, their multiphasic nature is less apparent than in our data, and the authors assign each collagen a single transition midpoint. Consistant with our observation they find that V, is significantly less stable than V,. Linsenmayer et al. (33) also show a transition curve which is not obviously multiphasic. On the other hand they show

4 The residue numbering begins with the first residue of the first amino-terminal tripeptide unit of the main triple helix. The 02(X1) chain was assumed to be the same length as the 01(X1).

by guest on March 13, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Thermal Stability of Type XI and V Collagens

12 3 4

T- 50

40

30 E

z z

20

10

0 60 90 120 150 160 210 240 270 300

LENGTH (nm)

- I * I

FIG. 7. SDS-polyacrylamide gel of type V, and type XI col- lagens digested with trypsin. A 510% polyacrylamide gel was used to separate the reduced peptides after digestion with trypsin at various temperatures. Lane I shows type V, collagen digested at 38 “C; lane 2, type V, collagen digested at 40 ‘C; lane 3, type XI collagen digested at 38 “C; and lane 4, type XI collagen digested at 40 “C. The arrows indicate the peptides that were sequenced and the numbers correspond to the sequences in Table II.

6 35 35

30

25

that the thermal destruction of antigenic determinants oc- curred at different temperatures for two different antibodies. Among the possible explanations offered was a greater than average stability of some portions of the molecule.

Type V and XI collagens are believed to be closely related to interstitial collagens types I, II, and III (11, 13-16, 34). The length of the triple helix is about 300 nm and there are no interruptions in the Gly-X-Y sequence. Both are believed to participate in heterotypic fibrils together with interstitial collagens (7-9). An evaluation of the thermal triple helix to coil transition shows some significant differences compared to other interstitial collagens. Although the other interstitial collagens follow a highly cooperative unfolding transition without observable intermediates these collagens unfold in several transitions with stable intermediates. For type V, these intermediates consist of triple helical regions which are 160 nm long at 38 “C and 126 nm at 40 “C. Sequence analysis of an intermediate fragment generated by trypsin digestion indicates that a stable 160-nm long stretch of triple helix starting at position 430 is flanked by regions of lower stability. At 40 “C this fragment is further reduced to a length of 126 nm, which is consistent with the 40% of the ORD signal observed at this temperature. The intermediates of type XI collagen are very similar to type V, collagen. Two fragments of about 135 nm in length are flanked by regions of lower stability at 38 “C and start at positions 495 and 519, respec- tively. At 40 “C the length is reduced to 113 nm with starting positions at residues 585 and 618, again consistent with the ORD signal at that temperature. Those results indicate that the more stable domain starts near the middle of the triple helix and extends close to the carboxyl-terminal end.

The basis for the domains of different stability is unclear. Proline residues and particularly hydroxyproline residues con- tribute significantly to the stability of the triple helix. An analysis of the amino acid composition of the constituent chains of types V and XI collagen reveals that there are no

LENGTH (nm)

C 60

f 50

40

iFI g 30

2

20

10

1 0 60 90 120 150 160 210 240 270 300

LENGTH (nm)

60

50

40

30

20

10

0 c,l I

length after incubation at 25 “C; stippled bars, 36 “C; black bars, 38 “C; and white bars, 40 “C, respectively, in all panels. Panel A, human type V, collagen (c&x2) with the following average length and standard deviation: 25 “C, 292 f 7 nm; 38 “C, 160 f 10 nm; and 40 “C, 126 + 20 nm. Panel B, human type V, collagen (nlot2a3), 25 “C, 286 f 6 nm; 36 “C, 157 + 14 nm; 38 “C, 155 f 17 nm. Panel C, bovine type XI collagen 25 “C!, 292 f 11 nm; 38 “C, 135 + 11 nm; and 40 “C, 113 f 17 nm.

FIG. 6. Length distribution of type V,, type V,, and type XI collagen after trypsin digestion at various temperatures as determined by rotary shadowing. The collagens were incubated for 10 min at the indicated temperature, digested for 2 min with trypsin, and prepared for electron microscopy. The length of the molecules was determined and the results plotted as number of molecules uersus the length (in lo-nm increments). Grey bars indicate

by guest on March 13, 2020

http://ww

w.jbc.org/

Dow

nloaded from

10086 Thermal Stability of Type XI and V Collagens

TABLE II

Amino acid sequences of intermediates obtained by digestion with trypsin at various temperatures The possible constituent chains and positions of the first residues are indicated. The human DNA-derived

sequences for the chains of type XI are given under the peptide sequences. Z denotes hydroxyproline.

Collagen Temperature Sequence Chain Position

Tme V, PeDtide 1 Peptide 2 Peptide 3

Type XI Peptide 4

38 TGPZGPZGVVGPQGPTGE 38 GAZGKDG?VGP?GP 40 GLZGPVGALGL?GNEGPZG

38

Peptide 5 38

Peptide 6 40

Peptide 7 40

“C

GLZGTAGGZGL?GN RGLPGTAGGPGLKGN

GLZGTZGTDGP?GA RGLPGTPGTDGPKGA

GLTGPIGPZG RGLTGPIGPPG

GAZGERGETGPZGPA RGAPGERGETGPZGPA

al(V)? aw 430 al(V)?

&2(X1) or2(XI) 495 a3(XI) cul(II) 519 a3(XI) al(II) 585 a3(XI) (Yl(I1) 618

significant differences in the amount of proline and hydrox- yproline (34, 35). Inspection of the amino acid sequences shows that these residues are not unusually clustered or distributed within the triple helical domains of types V and XI collagen when compared to types I and II (13-16, 36). With respect to other stab&zing forces, the total transition enthalpies of types V and XI collagen are very similar to types I and II collagen (3, 18).

Because the complexity of the thermal denaturation profiles of type V and XI collagens are due to regions of the triple helix which denature at a temperature lower than type II, one might argue that these coliagens have suffered some insult during extraction and purification, such as exposure to con- ditions of pH, ionic strength, chaotropic agent, or chromatog- raphy sufficient to cause partial denaturation. In these exper- iments, all of the collagens have been exposed to the same conditions, including pepsin concentration, urea concentra- tion, types of chromatography, and buffer solutes with the exception that type V collagen is exposed to 4 M NaCl. Type II collagen is not exposed to greater than 1 M NaCl, but neither is type XI collagen. We cannot rule out that types V and XI are disproportionately sensitive to these conditions compared to type II, since we cannot measure the stability of individual molecules in uiuo. However, this would still argue that there is some intrinsic difference in the stability of the triple helical conformation between these two groups of inter- stitial collagens.

The only appreciable difference in the amino acid sequences of types V and XI collagen compared to the other interstitial collagens is the small number of tripeptide units containing alanine. It has been shown that the tripeptide units Gly-Ala- Hyp and Gly-Pro-Ala can form stable triple helices in colla- gen-like polytripeptides and probably contribute significantly to the stability of collagens (37, 38). This might explain the regions of lower stability in types V and XI collagen since the distribution of these tripeptide units is biased towards the more stable regions.

It is known that the matrix forms of types V and XI collagen contain additional domains at the ends of the major triple helix (11, 12, 16, 32). Because we used pepsinized materials which lack these portions, we cannot exclude the possibility that the matrix forms are further stabilized by these regions. Such domains in the other interstitial collagens add little to stability, however.

At present we can only speculate about the physiological significance of the less stable regions in these molecules, particularly because the functions of types V and XI collagen

are unknown. Lower thermal stability might lead to greater flexibility and certainly to protease sensitivity. Based on the observations that these collagens are located within the het- erotypic collagen fibrils, one could hypothesize that they form a scaffolding which directs fibril formation. Enhanced flexi- bility could contribute to subtle changes in packing and pos- sibly to branch points. The triple helical domain is relatively resistant to proteolysis, requiring specific enzymes for turn- over in I&O. Vertebrate collagenase cleaves types I, II, and III, but not types V and XI (11,34,39). A matrix metallopro- teinase (type V collagenase/gelatinase) preferentially cleaves type V at 32 “C, leaving behind a portion of the molecule (40). At 37 “C such an enzyme would be expected to degrade greater than 60% of a type V molecule. During remodeling or resorp- tion, the less stable regions of type V and XI, which are protected by packing within the heterotypic fiber, would be exposed by the action of vertebrate collagenase and cleaved.

Acknowledgments--We wish to acknowledge Doug Keene for rotary shadowing and electron microscopy, Kerry Maddox and Dr. Rob Glanville for amino acid analysis and sequencing of the fragments, and Bob Poljak and Bruce Donaldson for assistance with preparation of tissues.

REFERENCES

1. Bornstein, P., and Traub, W. (1979) in Z%e Proteins (Neurath, H., Hill, R. R., and Breder, C. R., eds) Vol. 4, pp. 411-632, Academic Press, Orlando, FL

2. Fraser, R. D. B., MacRae, T. P., and Suzuki, E. (1979) J. Mol. Biol. 129,463-481

3. Privalov, P. L. (1982) Adu. Protein Chem. 35, l-104 4. Kivirikko, K., and Myllyll, R. (1980) in The Enzymology of Post-

translational Modification of Proteins (Freedman, R. B., and Hawkins, H. C., eds) Vol. 1, pp. 53-104, Academic Press, Orlando, FL

5. Burjanadze, T. V. (1979) Biopolymers 18.931-938 6. Sakakibara, S., Inouye, K., Shudo, K., Kishida, Y., Kobayashi,

Y., and Prockop, D. J. (1973) Biochim. Biophys. Acta 303, 198-202

7. Mendler, M., Eich-Bender, S. G., Vaughan, L., Winterhalter, K. H., and Bruckner, P. (1989) J. Celf Biot. 108, 191-197

8. Birk, D. E., Fitch, J. M., Babiarz, J. P., and Linsenmayer, T. F. (1988) J. Cell Biol. 106,999-1008

9. Keene, D. R., Sakai, L. Y., Blchinger, H. P., and Burgeson, R. E. (1987) J. Cell Biol. 105, 2393-2402

10. Kiihn, K. (1987) in The Structure and Function of Collagen Types (Mayne, R., and Burgeson, R. E., eds) pp. l-42, Academic Press, Orlando, FL

11. Fessler, J. H., and Fessler, L. I. (1987) in The Structure and Function of Collagen Types (Mayne, R., and Burgeson, R. E., eds) pp. 81-103, Academic Press, Orlando, FL

by guest on March 13, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Thermal Stability of Type XI and V Collagens

12.

13.

14.

15.

16.

17. 18.

19.

Engel, J. (1987) Adu. Meat Res. 4, 145-161 Bachinger, H. P., Morris, N. P., and Davis, J. M. (1989) in

Cytoskeletal and Extracellular Proteins (Aebi, U., and Engel, J., eds) Vol. 3, pp. 171-181, Springer-Verlag, Berlin

Davis. J. M.. Boswell. B. A.. and Bachineer. H. P. (1989) J. Biol.

20. Ch;m. 26&8956-8962

I

Niyibizi, C., Fietzek, P. P., and van der Rest, M. (1984) J. Biol. Chem. 259,14170-14174

21.

22.

Hashimoto, Y., Kobayashi, K., Hoshino, T., Aoyama, H., and Hayakawa, T. (1988) J. Inuest. Dermatol. 91, 238-242

Burgeson, R. E., and Hollister, D. W. (1979) Biochem. Biophys. Res. C’ommun. 87, 1124-1131

23. Morris, N. P., Keene, D. R., Glanville, R. W., Bentz, H., and

24. 25. 26.

Burgeson, R. E. (1986) J. Biol. Chem. 261, 5638-5644 Freire, E., and Biltonen, R. L. (1978) Biopolymers 17, 463-479 Freire, E., and Biltonen, R. L. (1978) Biopolymers 17,481-496 Matsudaira, P. (1987) J. Biol. Chem. 262, 10035-10038

Morris, N. P., and Bachinger, H. P. (1987) J. Biol. Chem. 262, 11345-11350

Bernard, M., Yoshioka, H., Rodriguez, E., van der Rest, M., Kimura, T., Ninomiya, Y., Olsen, B. R., and Ramirez, F. (1988) J. Biol. Chem. 263, 17159-17166

Kimura, T., Cheah, K. S. E., Chan, S. D. H., Lui, V. C. H., Mattei, M-G., van der Rest, M., Ono, K., Solomon, E., Ninomiya, Y., and Olsen, B. R. (1989) J. Biol. Chem. 264, 13910-13916

Weil, D., Bernard, M., Gargano, S., and Ramirez, F. (1987) Nucleic Acids Res. 15, 181-198

Woodbury, D., Benson-Chanda, V., and Ramirez, F. (1989) J. Biol. Chem. 264, 2735-2738

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37. 38.

39.

40.

Rhodes, R. K., and Miller, E. J. (1978) Biochemistry 17, 3442- 3448

Bentz. H., Bachinger, H. P., Glanville, R. W., and Kuhn, K. (1978) i&r. J. B&hem. 92,563-567

Sage, H., Pritzl, P., and Bornstein, P. (1981) Biochemistry 20, 3778-3784

Trelstad, R. L., Lawley, K. R., Hayashi, K., Ehrlich, H. P., and Silver, F. H. (1981) Collagen Relat. Res. 1, 39-52

Rhodes, R. K., and Miller, E. J. (1981) Collagen Relat. Res. 1, 337-343

Broek, D. L., Madri, J., Eikenberry, E. F., Brodsky, B. (1985) J. Biol. Chem. 260,555-562

Linsenmayer, T. F., Fitch, J. M., Schmid, T. M., Zak, N. B., Gibney, E., Sanderson, R. D., and Mayne, R. (1983) J. Cell Biol. 96, 124-132

Eyre, D., .and Wu, J-J. (1987) in Structure and Function of Collagen Types (Mayne, R., and Burgeson, R. E., eds) pp. 261- 281, Academic Press, Orlando, FL

Burgeson, R. E., Hebda, P. A., Morris, N. P., and Hollister, D. W. (1982) J. Biol. Chem. 257,7852-7856

Baldwin, C. T., Reginato, A. M., Smith, C., Jimenez, S. A., and Prockop, D. J. (1989) Biochem. J. 262,521-528

Dolz, R., and Heidemann, E. (1986) Biopolymers 25, 1069-1080 Thakur, S., Vadolas, D., Germann, H.-P., and Heidemann, E.

(1986) Biopolymers 25, 1081-1086 Eyre, D. R., Wu, J-J., and Woolley, D. E. (1984) Biochem.

Biophys. Res. Commun. 118, 724-729 Collier, I. E., Wilhelm, S. M., Eisen, A. Z., Marmer, B. L., Grant,

G. A., Seltzer, J. L., Kronberger, A., He, C., Bauer, E. A., and Goldberg, G. I. (1988) J. Biol. Chem. 263, 6579-6587

by guest on March 13, 2020

http://ww

w.jbc.org/

Dow

nloaded from

N P Morris, S L Watt, J M Davis and H P Bächinger3.

collagen and human type V collagens alpha 1(2) alpha 2 and alpha 1 alpha 2 alpha Unfolding intermediates in the triple helix to coil transition of bovine type XI

1990, 265:10081-10087.J. Biol. Chem. 

  http://www.jbc.org/content/265/17/10081Access the most updated version of this article at

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/265/17/10081.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at

by guest on March 13, 2020

http://ww

w.jbc.org/

Dow

nloaded from